Since you mention lights, one thing I hate about BMWs (at least in the US) is the blue headlights. If I see a harsh blue glare coming up behind me I know it is a BMW.
Better than being tailgated by an oversized pickup with headlights shining right into my side mirrors, though.
Shareholders are not on the line for a company's debts- shares can not go below zero dollars. I don't think you have a firm grasp on what the stock market's function is, I suggest for your own good that you learn- it is not a casino, it is a way for a company to raise capital.
If I go into a casino and place a bet, I might get a little extra, but I cannot lose more than I put down. I am not on the line for the casino's debts.
There are different rules surrounding stock, but you are still gambling that a particular business plan will work.
Even if MS is able to make Windows good at what it is and generally reliable, what it is is not a high-SLA platform intended for mission critical systems, so there's really no excuse. I don't think NSA/CIA/DoD would say, "The security model of Windows isn't quite as bad as the/. crowd likes to say it is. Sure, we haven't reviewed it, but the IT guy says it will help us leverage synergy to effect better ROI."
Identity is not provable. All that is provable, is possession of a token (or, multiple tokens, such as access to email address, telephone, an apparently valid photo id...) that supposedly establishes identity.
Identity is not provable? Depends on your definition of identity, I guess. I would say that identity is provable, at least in that, if I am given a public key when I walk into a bank to open an account, I can verify that the website I communicate with later is in fact owned by the same people I interacted with at the physical branch, which is all I care about. This is much better than the current certificate authority situation, in which I have to trust that a third party did their due diligence and are in fact certifying the people I think they are (and there have been cases where they are not, i.e. a phisher registers a legitimate-looking domain and gets a cert). The problems with identity these days is that it is too often based on poorly secured tokens, and too infrequently on identity established by face to face interaction.
I'm not surprised anymore at articles such as this one. Our DNA is basically a blue print of who we are. Our limitations, strengths, etc...
I think it's interesting the way people tend to polarize the physical vs emotional reasons for human behavior. It seems there is often a sort of threshold that separates them. If a person is mentally retarded, we pretty much take it for granted that there is some medical reason. If they're just not all that bright, though, then they or their parents must not have tried hard enough.
Hopefully as more and more research like this comes out, more people will acknowledge that 99% of the time, the reasons for a person being the way they are are somewhere between the two, and more complex than either. That much seems obvious to me, but it doesn't seem to occur to many people.
I think the polarization gets abused and clouds the real issues. To take obesity as an example, an obese patient might claim they have a medical problem they can't control, to try and make it more socially acceptable, whereas a doctor might claim it's a problem of free will and refuse to investigate non-behavioral treatments. Likely it is both, and we shouldn't just feel sorry for obese people because they can't control their actions, and thus remove social pressure to change their behavior, but neither should researchers simply proclaim that there is no environmental factor, when they could possibly overlook such tremendous public health gains.
Except that the attackers also chose not to disclose the vulnerability to the MTBA before giving the talk, and even refused to disclose their materials when the MTBA found out about the talk anyway.
And that there was apparently significant social engineering involved. Security should be robust when its implementation is widely known, but certain things like private keys have to remain secret. If I just con someone out of their sensitive information, it doesn't really mean a system that uses it is flawed.
And then, I'm not really sure why they went after this system in the first place. It's not a system whose flaws could endanger the public (as opposed to banking or military systems, for example). If it was a regular paper ticket system, we would just call someone who forges tickets a criminal, and a rather petty one at that.
So we have a group of people who used dubious methods to break into a system, who in doing so could not have helped anyone other than the owners of the system, and who chose not to disclose to them. Not exactly whitehat. I don't see any problem with calling these people criminals.
If you did want to solve the problem of account lockout, you could try this: the first time an incorrect password happens, lock the account for 0.1 seconds. For every subsequent attempt, increase the lockout time by 10. After 3 bad guesses, you'd have to wait almost 2 minutes. After four guesses: 16 minutes. Five guesses: 2+3/4 hours. Six guesses: a day and 3 hours. Seven guesses: a week and a half. Eight guesses: 3+1/2 months. So, on the one hand, if the account does get DOS'd, it's merely "relatively" DOS'd to some extent; on the other hand, if Evil Hacker really wanted to DOS the account to a great extent, then it would be inconvenient for Evil Hacker, who might actually wait 2 minutes for the fourth guess but probably won't wait 16 minutes to enter the fifth guess. The Innocent End User, checking her account at the end of the day, might not even know that it had been semi-DOS'd.
Lots of creative ways you can solve these problems. I came up with this in the time it took me to type this post. I'm sure others have more ideas.
A good high-entropy password will be essentially impossible to brute-force anyway. And your solution still lets an effectively permanent lockout happen if the attacker has a dedicated machine to do it with (though, if they did it for several weeks, they're more likely to get caught).
GP is correct that it's always a trade-off, in the sense that the allowed frequency of guesses is inversely proportional to the lock-out period. I think the problem is that people who use insanely bad passwords can't be protected except by draconian measures like an indefinite lockout. So I think the real solution is to train people in how to make and handle good passwords, but that's not always possible. If we can even get people to use moderately secure passwords in place of pet names or whatever, then a pretty reasonable rate limiting (say, three attempts every half hour) will stop brute-forcing.
(i.e. email addresses for IDs and short crackable passwords)
Are you seriously labelling email address as ID a security vulnerability? Usernames are not supposed to be sensitive data. That's what passwords are for.
I receive and pay a number of recurring bills online, and I find that the site that made my email address my userid the most sensible. The problem with making up your own ID is that if you try to reuse it across sites, half the time it will already be taken, whereas an email address is a globally unique identifier.
(I always found it BIZARRE that banks and its employees aren't trained to use PGP and the like for even large moneyed account holders and more sensitive information)
As long as my identity is known when I set up online access (say I do it at the branch when opening the account), a password authenticates me to the bank just as well as a private key can. And the bank is already authenticated to me using PKI through signed certs. The only advantage of the bank having a public key from me is so I can digitally sign, but that's only relevant for the bank to prove to someone else what I authorized, and so only matters if I don't trust the bank.
With the usual password system, I create a way to authenticate to one party. If someone gets my password, I can notify the bank and establish a new one. But with PKI, if I go around giving my public key to everyone, a compromise of my private key is like using the same password everywhere and then having it compromised. And if public keys become a de facto identifier used with third parties the way SSNs are now, changing your key becomes like changing your SSN and it's not going to be pleasant.
Given that I already have to trust the bank to have an account there, and that all the bank needs to know is that the person doing transactions is the same person who opened the account, using PKI for the account holder really adds nothing. In fact, if I was truly paranoid, I would want to remain anonymous to the bank, so I would prefer the existing system to PKI, and what I would actually want instead is just to make it two-factor.
Well now wait a minute. Housing prices across CA are still well above where they should be if you look at historical trends. Just because a bunch of speculators drive up the price of something, does that mean you should get paid more so that you don't feel the effects?
Want to work in Silicon Valley? Odds are you'll be renting, even if you work for Google, unless you're just stupid with money.
In addition to the other points people bring up, just moving from an ICE to an electric motor also creates a large increase in efficiency. EVs will force the development of better motor and battery technology which will be able to directly power a car, at a reasonable cost and with a reasonable lifespan. Current hybrids have small batteries and weak motors and generally power the wheels directly from the ICE, but using only electric torque with an ICE generator could greatly increase the mileage, even without the benefits of grid power. Diesel locomotives have been built this way for decades.
"6% of users generate 50% of clicks" is a much different statistic than a 6% click-through rate. 6% is very high for online advertising; 1-2% is a reasonable range. What the summary claims (not that the summary is necessarily accurate) is that the demographics who respond to the ads are not necessarily generating revenue for the advertisers, which if true has much worse implications for online ads than a low click-through rate.
That's just click spam, and really, it's a stupid thing to do. Online advertising providers make an effort to filter out clicks like yours, so it's possible your action doesn't actually net the site any revenue. More importantly, it only serves to dilute the effectiveness of online advertising and hurt revenue. Particularly if the ads you're seeing are poorly targeted, it would be better to not click on them, so that the advertising provider and the advertiser can learn what targeting works and increase the effectiveness of the online ad market, which is ultimately better for the site you're visiting.
Besides, it's not, literally, your 0.02, it's the advertiser's. If you really want to support the site but you're not interested in the ads, you might try directly sending a contribution, rather than trying to cheat it out of someone else.
How about really giving customers unlimited bandwidth? If they lack the infrastructure to support what they claim, then they should get better lines.
This statement is utterly stupid. It is harder to develop backbone capacity than last mile capacity, and ISPs have a very limited amount of backbone capacity. If they can supply a 10M last mile to 1000 customers and only have 1G of backbone, it still makes a lot more sense to give everyone a 10M line than to give everyone a 1M line, because not everyone's going to use it at once and this allows a lot more efficient allocation of bandwidth to whoever's demanding it at any given time. I think that in some instances they could do a better job of this allocation, but this is exactly what they are trying to do with a market solution, and it's no reason to choke off everyone's last mile.
Even if the technology was available to give ISPs a blazingly fast cheap backbone that would let everyone saturate existing last-mile technology, in such a case it would be likely that better last-mile technology exists as well, and you run into the same problem. If you're really so concerned about being able to saturate your line 24 hours a day, you can get a line with a higher SLA (and pay the true market value of the bandwidth). Alternatively, you could exercise some courtesy and just not leave BitTorrent downloading 24/7.
You're right that no engineer in his right mind would willingly design such a thing, but 1) there's lots of bad engineers out there, and 2) the engineers don't make these decisions, management does. If management says, "design a car using electronic brakes and steering, with no mechanical backup, because it'll save us money", the engineer either does it or finds a new job. Engineers are just peons.
Do you have an example of a DbW vehicle which does not have mechanical steering in case of failure, and/or does not have a mechanical brake system (keeping in mind that, if the primary brakes are electronic and fail, there's still the parking brake)?
Do you have evidence that DbW is cheaper than mechanical and that any auto manufacturers are looking at DbW as a cost-saving option?
'd be willing to bet that the first thing that happens is that an officer jumps the gun a little and uses it during, say, a routine traffic stop, causing an accident (as in the worst-case scenario described) and an ensuing lawsuit. Then it's back to the drawing board for a new crazy idea.
Since when are cops accountable for reckless actions such as this? Many departments have yet to figure out that tasing someone is use of force.
Or, you know, you can just clock your emulator so it runs at the actual hardware speed. I always did hate that instantaneous game over I get on NES emulators...
If gasoline makes up 50% of our oil useage and 10% of Americans using public transportation means cutting oil consumption by 40%, that means that we'll see a 80% reduction in gasoline useage from a 10% increase in transportation.
Actually, the comment was that it would cut imports by 40%. Supposing cutting usage by 10% (which also is not the same thing that was mentioned, but we'll ignore that for now) means 40% less imports, that means only that the U.S. satisfies 75% of its oil demand domestically.
Regardless of whether the IT sector is _technically_ in the wrong it's commonly accepted that in this area we work with powers of two. The fact that people have to explicitly explain this fact shows that everyone expects it to be that way. The HDD manufacturers damn well know this and fairly blantantly use measurements which would commonly be interpreted more favourably.
At the same time, this has been going on with hard drives for over a decade. I just always expect capacity to be 10-based and not 2-based when I buy a drive. It's a stupid system, but if everyone is doing it and competing with each other, whatever the market will pay me per 10^9 bytes is close to the actual market value of 10^9 bytes, and not 2^30 bytes. While I'm glad it's going to change back, it seems unfair that Seagate is taking a blow for this if it's not everyone, and the appropriate time for such a lawsuit would have been back when a few manufacturers started this practice and forced everyone to follow, not long after it's established, because at this point manufacturers can no longer use it as a way to inflate their prices.
Of course, there's also the problem that the mismatch grows with the units - 1KB is only off by 2.4%, while 1GB is off by 7.4%. I suspect this is why it took so long for anyone to raise a stink about it.
Nowhere does TFA say Google is trying to cut your bill by showing you ads on your phone, or anything like that. It talks about "handsets tailored to its new mobile-phone operating system," and adds, "The phones also would be open to third-party application development, potentially spurring development of new features." It sounds like all they are trying to do is open up the hardware. Verizon still owns the network and can sell you whatever data plan they want. The motivation is not hard to spot - if Google is successful with this, their apps (gmail, maps, etc) would be available to a lot more wireless devices, since carriers currently try to lock out access to a lot of third-party apps, and lock developers out of the hardware. Google doesn't need to charge anybody anything for doing this; their reward is increased market share as more people (choose freely to) use Google services on their phones.
If you actually paid attention what's going on, Google is speculated to be making an OS for phones, not phones themselves, and certainly not the network the phones will connect to. This is good - if wireless companies actually adopt this, it means more interoperability and less lock-down, and an open platform for you to make your own phone apps, etc. How you get out of this that Google will be tracking your phone calls is beyond me, when no data from a Google phone would even go through Google unless Google is where you were browsing on it.
Google has come a long way from the benign trustworthy startup they were many years ago, but please, base your bashing on actual events.
Um, no, the problem is there were no proof reviewers. This was just Wolfram being overly eager for a chance to once again assert his superiority over the scientific community.
"As far as I know, no member of the committee has passed on the
validity of this 40 page proof. The determination that Smith's proof
is correct seems to have been made entirely by the Wolfram
organization. My understanding is that the I/O involves complex encodings." http://cs.nyu.edu/pipermail/fom/2007-October/012132.html
My thought is this, a sinewave consists of 360 degress, correct? So if we are looking for an "acceptable" minimum number of sample points to define a "smooth" sine wave, 360 sample points is a good logical number to shoot for.
That makes no sense. 360 is just an arbitrary number of sections to break a periodic phenomenon into. You could also say it has 2*pi radians, or 1600 BAMs (or whatever that unit was the Doom engine used). It has no mathematical relation to how many samples are needed.
I think 4 or 5x the highest expected input frequency is plenty. Keep in mind that a good DA converter will produce smooth output and not just do linear interpolation or sample and hold, so really, what the data looks like is irrelevant; it's how much information is available to the DA converter and how well it can use that to reproduce the original signal.
Right now vinyl does a better job than current digital audio because it does a better job of representing the signal in motion.
As you said, analog and digital are both imperfect and it's about the amount of information from the original signal that is preserved. But I disagree that vinyl holds more information - I would argue it holds significantly less than CDs. The reason it sounds better to a lot of people I think has to do with the nature of the noise added to the signal, not the amount. Analog noise tends to be white noise and smooth; digital noise tends to be discrete. Think of the unpleasant artifacts you get in highly compressed mp3, for example - not really the same thing as sampling error, but it demonstrates the irritating nature of digital artifacts.
Looks like I'm wrong myself. But the point remains that the Nyquist theorem does not mean a CD accurately reproduces a 22kHz-limited source signal. The Nyquist theorem deals with an infinite number of samples, and so has only limited application to real-world sampling anyway. A finite digital sampling will still avoid aliasing below the Nyquist frequency, though.
Another reason for vinyl's sonic superiority is that no matter how high a sampling rate is, it can never contain all of the data present in an analog groove, Nyquist's theorem to the contrary.
The mathematics behind sampling theory is widely misunderstood, and unfortunately the author has fallen into the same trap. I would like just once to see someone properly reference the Nyquist theorem when debating the merits of different recording formats.
The Nyquist theorem is about aliasing, a phenomenon where a sampled wave comes out as a different frequency than the input wave, and this will happen any time the input wave is above half the sample rate, or Nyquist frequency. Nyquist's theorem states it will not happen below that frequency, and it's pretty intuitive - suppose you are sampling a pure frequency at at least twice the frequency; then you cannot jump over any contiguous positive or negative portion of the input, and so you can't get aliasing.
The Nyquist theorem is not about accurate reproduction. You can still sample the Nyquist frequency at the zero every time.
In addition, the "information content" of analog is irrelevent - first of all, no analog medium has "infinite information", due to quantum uncertainty. Second, even if it did, there's no such thing as a perfect analog recording, and what's important is the deviation from the source, not the amount of information. In fact, this sounds like an argument for digital, because with a high enough sample rate and small enough quantization, a digital signal is to our ears indistinguishable from the source, and has the added benefit of being able to be copied perfectly.
Slashdot Mirror
Actually I would assume that they mean 1/100 the temperature of space, on an absolute zero based scale.
Since you mention lights, one thing I hate about BMWs (at least in the US) is the blue headlights. If I see a harsh blue glare coming up behind me I know it is a BMW.
Better than being tailgated by an oversized pickup with headlights shining right into my side mirrors, though.
Shareholders are not on the line for a company's debts- shares can not go below zero dollars. I don't think you have a firm grasp on what the stock market's function is, I suggest for your own good that you learn- it is not a casino, it is a way for a company to raise capital.
If I go into a casino and place a bet, I might get a little extra, but I cannot lose more than I put down. I am not on the line for the casino's debts.
There are different rules surrounding stock, but you are still gambling that a particular business plan will work.
Even if MS is able to make Windows good at what it is and generally reliable, what it is is not a high-SLA platform intended for mission critical systems, so there's really no excuse. I don't think NSA/CIA/DoD would say, "The security model of Windows isn't quite as bad as the /. crowd likes to say it is. Sure, we haven't reviewed it, but the IT guy says it will help us leverage synergy to effect better ROI."
Identity is not provable? Depends on your definition of identity, I guess. I would say that identity is provable, at least in that, if I am given a public key when I walk into a bank to open an account, I can verify that the website I communicate with later is in fact owned by the same people I interacted with at the physical branch, which is all I care about. This is much better than the current certificate authority situation, in which I have to trust that a third party did their due diligence and are in fact certifying the people I think they are (and there have been cases where they are not, i.e. a phisher registers a legitimate-looking domain and gets a cert). The problems with identity these days is that it is too often based on poorly secured tokens, and too infrequently on identity established by face to face interaction.
I think it's interesting the way people tend to polarize the physical vs emotional reasons for human behavior. It seems there is often a sort of threshold that separates them. If a person is mentally retarded, we pretty much take it for granted that there is some medical reason. If they're just not all that bright, though, then they or their parents must not have tried hard enough.
Hopefully as more and more research like this comes out, more people will acknowledge that 99% of the time, the reasons for a person being the way they are are somewhere between the two, and more complex than either. That much seems obvious to me, but it doesn't seem to occur to many people.
I think the polarization gets abused and clouds the real issues. To take obesity as an example, an obese patient might claim they have a medical problem they can't control, to try and make it more socially acceptable, whereas a doctor might claim it's a problem of free will and refuse to investigate non-behavioral treatments. Likely it is both, and we shouldn't just feel sorry for obese people because they can't control their actions, and thus remove social pressure to change their behavior, but neither should researchers simply proclaim that there is no environmental factor, when they could possibly overlook such tremendous public health gains.
Except that the attackers also chose not to disclose the vulnerability to the MTBA before giving the talk, and even refused to disclose their materials when the MTBA found out about the talk anyway.
And that there was apparently significant social engineering involved. Security should be robust when its implementation is widely known, but certain things like private keys have to remain secret. If I just con someone out of their sensitive information, it doesn't really mean a system that uses it is flawed.
And then, I'm not really sure why they went after this system in the first place. It's not a system whose flaws could endanger the public (as opposed to banking or military systems, for example). If it was a regular paper ticket system, we would just call someone who forges tickets a criminal, and a rather petty one at that. So we have a group of people who used dubious methods to break into a system, who in doing so could not have helped anyone other than the owners of the system, and who chose not to disclose to them. Not exactly whitehat. I don't see any problem with calling these people criminals.
If you did want to solve the problem of account lockout, you could try this: the first time an incorrect password happens, lock the account for 0.1 seconds. For every subsequent attempt, increase the lockout time by 10. After 3 bad guesses, you'd have to wait almost 2 minutes. After four guesses: 16 minutes. Five guesses: 2+3/4 hours. Six guesses: a day and 3 hours. Seven guesses: a week and a half. Eight guesses: 3+1/2 months. So, on the one hand, if the account does get DOS'd, it's merely "relatively" DOS'd to some extent; on the other hand, if Evil Hacker really wanted to DOS the account to a great extent, then it would be inconvenient for Evil Hacker, who might actually wait 2 minutes for the fourth guess but probably won't wait 16 minutes to enter the fifth guess. The Innocent End User, checking her account at the end of the day, might not even know that it had been semi-DOS'd.
Lots of creative ways you can solve these problems. I came up with this in the time it took me to type this post. I'm sure others have more ideas.
A good high-entropy password will be essentially impossible to brute-force anyway. And your solution still lets an effectively permanent lockout happen if the attacker has a dedicated machine to do it with (though, if they did it for several weeks, they're more likely to get caught).
GP is correct that it's always a trade-off, in the sense that the allowed frequency of guesses is inversely proportional to the lock-out period. I think the problem is that people who use insanely bad passwords can't be protected except by draconian measures like an indefinite lockout. So I think the real solution is to train people in how to make and handle good passwords, but that's not always possible. If we can even get people to use moderately secure passwords in place of pet names or whatever, then a pretty reasonable rate limiting (say, three attempts every half hour) will stop brute-forcing.
Inadequate policies for user ids and passwords
(i.e. email addresses for IDs and short crackable passwords)
Are you seriously labelling email address as ID a security vulnerability? Usernames are not supposed to be sensitive data. That's what passwords are for.
I receive and pay a number of recurring bills online, and I find that the site that made my email address my userid the most sensible. The problem with making up your own ID is that if you try to reuse it across sites, half the time it will already be taken, whereas an email address is a globally unique identifier.
(I always found it BIZARRE that banks and its employees aren't trained to use PGP and the like for even large moneyed account holders and more sensitive information)
As long as my identity is known when I set up online access (say I do it at the branch when opening the account), a password authenticates me to the bank just as well as a private key can. And the bank is already authenticated to me using PKI through signed certs. The only advantage of the bank having a public key from me is so I can digitally sign, but that's only relevant for the bank to prove to someone else what I authorized, and so only matters if I don't trust the bank.
With the usual password system, I create a way to authenticate to one party. If someone gets my password, I can notify the bank and establish a new one. But with PKI, if I go around giving my public key to everyone, a compromise of my private key is like using the same password everywhere and then having it compromised. And if public keys become a de facto identifier used with third parties the way SSNs are now, changing your key becomes like changing your SSN and it's not going to be pleasant.
Given that I already have to trust the bank to have an account there, and that all the bank needs to know is that the person doing transactions is the same person who opened the account, using PKI for the account holder really adds nothing. In fact, if I was truly paranoid, I would want to remain anonymous to the bank, so I would prefer the existing system to PKI, and what I would actually want instead is just to make it two-factor.
Want to work in Silicon Valley? Odds are you'll be renting, even if you work for Google, unless you're just stupid with money.
In addition to the other points people bring up, just moving from an ICE to an electric motor also creates a large increase in efficiency. EVs will force the development of better motor and battery technology which will be able to directly power a car, at a reasonable cost and with a reasonable lifespan. Current hybrids have small batteries and weak motors and generally power the wheels directly from the ICE, but using only electric torque with an ICE generator could greatly increase the mileage, even without the benefits of grid power. Diesel locomotives have been built this way for decades.
"6% of users generate 50% of clicks" is a much different statistic than a 6% click-through rate. 6% is very high for online advertising; 1-2% is a reasonable range. What the summary claims (not that the summary is necessarily accurate) is that the demographics who respond to the ads are not necessarily generating revenue for the advertisers, which if true has much worse implications for online ads than a low click-through rate.
That's just click spam, and really, it's a stupid thing to do. Online advertising providers make an effort to filter out clicks like yours, so it's possible your action doesn't actually net the site any revenue. More importantly, it only serves to dilute the effectiveness of online advertising and hurt revenue. Particularly if the ads you're seeing are poorly targeted, it would be better to not click on them, so that the advertising provider and the advertiser can learn what targeting works and increase the effectiveness of the online ad market, which is ultimately better for the site you're visiting.
Besides, it's not, literally, your 0.02, it's the advertiser's. If you really want to support the site but you're not interested in the ads, you might try directly sending a contribution, rather than trying to cheat it out of someone else.
How about really giving customers unlimited bandwidth? If they lack the infrastructure to support what they claim, then they should get better lines.
This statement is utterly stupid. It is harder to develop backbone capacity than last mile capacity, and ISPs have a very limited amount of backbone capacity. If they can supply a 10M last mile to 1000 customers and only have 1G of backbone, it still makes a lot more sense to give everyone a 10M line than to give everyone a 1M line, because not everyone's going to use it at once and this allows a lot more efficient allocation of bandwidth to whoever's demanding it at any given time. I think that in some instances they could do a better job of this allocation, but this is exactly what they are trying to do with a market solution, and it's no reason to choke off everyone's last mile.
Even if the technology was available to give ISPs a blazingly fast cheap backbone that would let everyone saturate existing last-mile technology, in such a case it would be likely that better last-mile technology exists as well, and you run into the same problem. If you're really so concerned about being able to saturate your line 24 hours a day, you can get a line with a higher SLA (and pay the true market value of the bandwidth). Alternatively, you could exercise some courtesy and just not leave BitTorrent downloading 24/7.
You're right that no engineer in his right mind would willingly design such a thing, but 1) there's lots of bad engineers out there, and 2) the engineers don't make these decisions, management does. If management says, "design a car using electronic brakes and steering, with no mechanical backup, because it'll save us money", the engineer either does it or finds a new job. Engineers are just peons.
Do you have an example of a DbW vehicle which does not have mechanical steering in case of failure, and/or does not have a mechanical brake system (keeping in mind that, if the primary brakes are electronic and fail, there's still the parking brake)?
Do you have evidence that DbW is cheaper than mechanical and that any auto manufacturers are looking at DbW as a cost-saving option?
'd be willing to bet that the first thing that happens is that an officer jumps the gun a little and uses it during, say, a routine traffic stop, causing an accident (as in the worst-case scenario described) and an ensuing lawsuit. Then it's back to the drawing board for a new crazy idea.
Since when are cops accountable for reckless actions such as this? Many departments have yet to figure out that tasing someone is use of force.
Or, you know, you can just clock your emulator so it runs at the actual hardware speed. I always did hate that instantaneous game over I get on NES emulators...
If gasoline makes up 50% of our oil useage and 10% of Americans using public transportation means cutting oil consumption by 40%, that means that we'll see a 80% reduction in gasoline useage from a 10% increase in transportation.
Actually, the comment was that it would cut imports by 40%. Supposing cutting usage by 10% (which also is not the same thing that was mentioned, but we'll ignore that for now) means 40% less imports, that means only that the U.S. satisfies 75% of its oil demand domestically.
Regardless of whether the IT sector is _technically_ in the wrong it's commonly accepted that in this area we work with powers of two. The fact that people have to explicitly explain this fact shows that everyone expects it to be that way. The HDD manufacturers damn well know this and fairly blantantly use measurements which would commonly be interpreted more favourably.
At the same time, this has been going on with hard drives for over a decade. I just always expect capacity to be 10-based and not 2-based when I buy a drive. It's a stupid system, but if everyone is doing it and competing with each other, whatever the market will pay me per 10^9 bytes is close to the actual market value of 10^9 bytes, and not 2^30 bytes. While I'm glad it's going to change back, it seems unfair that Seagate is taking a blow for this if it's not everyone, and the appropriate time for such a lawsuit would have been back when a few manufacturers started this practice and forced everyone to follow, not long after it's established, because at this point manufacturers can no longer use it as a way to inflate their prices.
Of course, there's also the problem that the mismatch grows with the units - 1KB is only off by 2.4%, while 1GB is off by 7.4%. I suspect this is why it took so long for anyone to raise a stink about it.
Nowhere does TFA say Google is trying to cut your bill by showing you ads on your phone, or anything like that. It talks about "handsets tailored to its new mobile-phone operating system," and adds, "The phones also would be open to third-party application development, potentially spurring development of new features." It sounds like all they are trying to do is open up the hardware. Verizon still owns the network and can sell you whatever data plan they want. The motivation is not hard to spot - if Google is successful with this, their apps (gmail, maps, etc) would be available to a lot more wireless devices, since carriers currently try to lock out access to a lot of third-party apps, and lock developers out of the hardware. Google doesn't need to charge anybody anything for doing this; their reward is increased market share as more people (choose freely to) use Google services on their phones.
If you actually paid attention what's going on, Google is speculated to be making an OS for phones, not phones themselves, and certainly not the network the phones will connect to. This is good - if wireless companies actually adopt this, it means more interoperability and less lock-down, and an open platform for you to make your own phone apps, etc. How you get out of this that Google will be tracking your phone calls is beyond me, when no data from a Google phone would even go through Google unless Google is where you were browsing on it.
Google has come a long way from the benign trustworthy startup they were many years ago, but please, base your bashing on actual events.
Um, no, the problem is there were no proof reviewers. This was just Wolfram being overly eager for a chance to once again assert his superiority over the scientific community.
"We're excited to announce that the $25,000 Wolfram 2,3 Turing Machine Research Prize has been won." http://cs.nyu.edu/pipermail/fom/2007-October/012120.html
"As far as I know, no member of the committee has passed on the validity of this 40 page proof. The determination that Smith's proof is correct seems to have been made entirely by the Wolfram organization. My understanding is that the I/O involves complex encodings." http://cs.nyu.edu/pipermail/fom/2007-October/012132.html
My thought is this, a sinewave consists of 360 degress, correct? So if we are looking for an "acceptable" minimum number of sample points to define a "smooth" sine wave, 360 sample points is a good logical number to shoot for.
That makes no sense. 360 is just an arbitrary number of sections to break a periodic phenomenon into. You could also say it has 2*pi radians, or 1600 BAMs (or whatever that unit was the Doom engine used). It has no mathematical relation to how many samples are needed.
I think 4 or 5x the highest expected input frequency is plenty. Keep in mind that a good DA converter will produce smooth output and not just do linear interpolation or sample and hold, so really, what the data looks like is irrelevant; it's how much information is available to the DA converter and how well it can use that to reproduce the original signal.
Right now vinyl does a better job than current digital audio because it does a better job of representing the signal in motion.
As you said, analog and digital are both imperfect and it's about the amount of information from the original signal that is preserved. But I disagree that vinyl holds more information - I would argue it holds significantly less than CDs. The reason it sounds better to a lot of people I think has to do with the nature of the noise added to the signal, not the amount. Analog noise tends to be white noise and smooth; digital noise tends to be discrete. Think of the unpleasant artifacts you get in highly compressed mp3, for example - not really the same thing as sampling error, but it demonstrates the irritating nature of digital artifacts.
Looks like I'm wrong myself. But the point remains that the Nyquist theorem does not mean a CD accurately reproduces a 22kHz-limited source signal. The Nyquist theorem deals with an infinite number of samples, and so has only limited application to real-world sampling anyway. A finite digital sampling will still avoid aliasing below the Nyquist frequency, though.
Another reason for vinyl's sonic superiority is that no matter how high a sampling rate is, it can never contain all of the data present in an analog groove, Nyquist's theorem to the contrary.
The mathematics behind sampling theory is widely misunderstood, and unfortunately the author has fallen into the same trap. I would like just once to see someone properly reference the Nyquist theorem when debating the merits of different recording formats.
The Nyquist theorem is about aliasing, a phenomenon where a sampled wave comes out as a different frequency than the input wave, and this will happen any time the input wave is above half the sample rate, or Nyquist frequency. Nyquist's theorem states it will not happen below that frequency, and it's pretty intuitive - suppose you are sampling a pure frequency at at least twice the frequency; then you cannot jump over any contiguous positive or negative portion of the input, and so you can't get aliasing.
The Nyquist theorem is not about accurate reproduction. You can still sample the Nyquist frequency at the zero every time.
In addition, the "information content" of analog is irrelevent - first of all, no analog medium has "infinite information", due to quantum uncertainty. Second, even if it did, there's no such thing as a perfect analog recording, and what's important is the deviation from the source, not the amount of information. In fact, this sounds like an argument for digital, because with a high enough sample rate and small enough quantization, a digital signal is to our ears indistinguishable from the source, and has the added benefit of being able to be copied perfectly.